MONETARY POLICY REPORT | SUMMER 2023 (box 1)
(authors: Ondřej Michálek, Matěj Šarboch)
Czech households have faced multiple extraordinary economic shocks in recent years. The coronavirus pandemic and the energy crisis were followed by a still ongoing period of very rapid growth in consumer prices. In response to these shocks, which are giving rise to great uncertainty, households are increasing their propensity to save. The saving and consumption behaviour of households directly affects domestic demand and hence plays an important role in overall macroeconomic developments. It is therefore also relevant to the monetary policy decisions of the central bank, as an elevated saving rate means a lower propensity of households to consume. Other things being equal, this implies lower demand pressures and more subdued inflation. By contrast, a sharp drop in the saving rate would mean growth in consumer appetite and upward pressure on inflation.
This box describes the motives behind the current surge in saving, the savings structure by income groups of the population, and the factors that can lead the saving rate to return close to its long-run average.
The CZSO defines the saving rate as the ratio of gross savings to gross disposable income. Savings comprise the unused part of gross disposable income. Gross disposable income is obtained by subtracting households’ current expenditure, such as taxes and social contributions, from their current income, the largest items of which are wages and salaries, entrepreneurs’ profits and social benefits. All these items, including savings, are flow (not stock) variables, hence they represent households’ nominal flows in the given quarter. The savings themselves may not be equal to the change in the stock of households’ bank deposits, as they may be converted into physical (tangible) investments, such as house purchases, and financial investments.
Since the start of 2020, the saving rate has been well above the average for 2012–2019, when it fluctuated around 12% (see Chart 1). Administrative constraints on households’ ability to spend during the first pandemic lockdown in spring 2020 caused the saving rate to rise sharply. The forced decline in households’ (nominal) consumption ultimately even outweighed the temporary drop in their income in spring 2020 (see Chart 2). Consumption recovered slightly during the subsequent easing of pandemic measures in the summer, but another wave of anti-pandemic measures introduced in the autumn and winter limited the operation of some shops and services again, narrowing households’ ability to consume. By then, the saving rate had reached an all-time high of 23%. In the course of 2021, the anti-pandemic measures were relaxed, households gradually became less cautious, their consumption recovered in both nominal and real terms and the saving rate started to fall (see Chart 3). In mid-2021, however, the decline in the saving rate started to be slowed by other (residual) categories, most notably worsening sentiment due to the energy crisis, growth in inflation above the tolerance band around the central bank’s target and increases in the CNB’s interest rates. All these factors intensified in 2022. The trend in the saving rate thus reversed in mid-2022 and the rate started to go up again. This mainly reflected increased saving triggered by caution, concerns about the energy crisis and a rapid general rise in consumer prices.
Chart 1 – The saving rate has been far above its long-run (pre-pandemic) average in recent years
seasonally adjusted saving rate in %; extent of revisions in pp
Note: The revisions are explicit from 2017 Q3 onwards, when the CZSO started to publish the seasonally adjusted saving rate. Until then, they mostly stemmed from seasonal adjustment.
Chart 2 – The saving rate remains elevated despite brisk growth in nominal household consumption
decomposition of saving rate; changes compared to 2019 Q4 and contributions in pp
Chart 3 – The saving rate has been higher in recent quarters due to the precautionary motive and the CNB’s monetary policy (high interest rates)
decomposition of saving rate; q-o-q changes and contributions in pp
Note: The decomposition of the saving rate into factors is based on a regression model derived from an ECB box (external link) in a specification where the saving rate depends on the unemployment rate (precautionary motive), the Stringency Index (anti-pandemic measures), interest rates and its previous level (other).
It is interesting to look at the distribution of savings by income quintile (see Chart 4). Until 2019, the four lowest income quintiles together accounted for just 10%–20% of total savings, with the poorest (first) quintile not saving at all. The wealthiest (fifth) quintile created all the rest of the savings, with a saving rate of just below 20%. In 2020, when most shops and services were locked down, the overall saving rate and the saving rate of the four lowest income quintiles both rose dramatically. The wealthiest quintile meanwhile kept saving at roughly the same rate. The savings of the four lowest income quintiles were thus partly forced, and the households concerned started to return to their usual consumer behaviour after shops and services reopened. We estimate that deferred consumption in this period amounted to around CZK 100 billion at constant prices.
Chart 4 – High-income households are maintaining the highest saving rate; it did not change much even during the Covid lockdowns
saving rate by income quintile in %
A similar situation could arise after the current elevated saving rate episode ends. Between mid-2022 and 2023 Q1, households’ savings at current prices were CZK 175 billion higher than would be consistent with the normal pre-pandemic level. If households were hypothetically to release this amount of money still this year, it would represent around CZK 120 billion at constant prices. This could foster growth in household consumption of 5 pp. However, we can assume that the situation this year is different from that in 2020 and that savings are being created mostly by high-income households, who are far less constrained by high prices of energy and other goods and services. In 2022, moreover, their incomes grew faster than those of low-income households, so they were able to put aside a larger proportion of their earnings for precautionary motives. Low-income households, who are now restricted the most in their real consumption by the high inflation and are likely also to have the greatest propensity to consume in the long run, are currently creating very low savings according to the available data. The future release of accumulated savings may therefore not be as huge as it was after the Covid measures were lifted.
The saving rate can be expected to decline gradually to its long-run average of around 12% over the next two years. This decline will be driven by the behaviour of the wealthiest households. It will be due mainly to improving consumer sentiment after the energy crisis fades away and inflation falls to low levels. This will be linked with renewed growth in real household consumption amid slowing nominal income growth and a gradual decrease in interest rates from their highest levels in more than 20 years. However, considerable uncertainty surrounds the pace of this decline in the saving rate.
 Put simply, the saving rate is the percentage of net (quarterly or annual) income that households set aside. The exact definition says that savings are obtained by subtracting household consumption from the sum of gross disposable income and the change in the net equity of households in pension funds reserves (transaction D.8). However, the latter item is equal to only around 1% of total gross disposable income.
 The great uncertainty regarding the saving rate is also illustrated by the considerable revisions made to it. During the pandemic, the revisions amounted to almost 4.5 pp.
 Households’ caution in the model in Chart 3 is proxied by the unemployment rate and thus represents households’ concerns about potential loss of employment and a drop in income.
 We use households’ satellite accounts from CZSO (external link) experimental statistics. The data are available for 2015–2020. They contain net disposable income and its components, final consumption expenditure and net savings broken down by income quintile.
 The total figure is not exactly equal to the national accounts data, as the saving rate in Chart 4 is calculated as the ratio of net savings to net disposable income. Unlike gross disposable income, net disposable income is adjusted for consumption of fixed capital (depreciation). This may explain why people with the lowest income are able to constantly create negative savings, i.e. consume more than they earn in each period.
 The calculation method, including the initial estimate of deferred consumption, was described in the box The pandemic has dramatically increased the household saving rate in MPR – Spring 2021.
 Evidence of growth in earnings last year based on the income distribution is provided in the full-year AEIS data (external link, available in Czech only).